This paper looks at some technologies that were developed at the University of Chicago to see how well they will work as the basis for the vision aspects of our new project, the Intelligent Classroom. The paper starts with a look at Gargoyle our real-time active vision framework and examines the features that make it a valuable resource to vision programers. It then looks at how Gargoyle has been tested and used so far. This experiment involved porting the Perseus real-time architecture's pointing task. Finally, the paper examines the Intelligent Classroom and shows the utility of Gargoyle by building on the results of the Perseus experiment. This paper also argues that for vision to be effective in the long term, it must be tightly coupled with a higher level planning or execution system.